Key words

1. Pancreatitis
2. Pseudocyst
3. WON
4. EUS
5. LAMS

Abbreviations

WON – walled off necrosis
EUS – Endoscopic ultrasound
LAMS – lumen apposing metal stents

Learning points

1. Pancreatic fluid collection are categorised according to the duration of the collection & the presence of necrosis (modified Atlanta classification)
2. Asymptomatic pancreatic collection does not require any intervention regardless of the size as most will remain asymptomatic or resolve without treatment
3. EUS guided pancreatic drainage is treatment of choice when technically feasible and should be delayed to 4 weeks from the onset of pancreatitis if clinically indicated.

Pancreatic fluid collections are a local complication following an attack of acute pancreatitis. This is secondary to release of proteolytic pancreatic fluid into the adjacent peritoneal cavity. Depending on the presence of necrosis and the age of the collection the Atllanta classification1 recognizes four types of pancreatic fluid collection which are summarized in table 1. Pseudocysts occur in around 5- 15 %² of case with pancreatitis while pancreatic necrosis in around 10-20 %^{1, 3.}

Infection of the necrotic collection can occur in 16-47% ^1,3and carries a mortality rate of 20-30%⁴. The presence & the characteristic of the pancreatic fluid collection can be confirmed with cross sectional imaging preferably contrast-enhanced computed tomography (CT) or MRI pancreas. CT is mostly used given its wider availability and good quality images. MRI pancreas lacks ionizing radiation and provide better soft tissue images and characterization of disconnected pancreatic duct but is expensive and time consuming ⁵. Imaging can be performed on admission with pancreatitis, when there is a clinical indication and up to or after the 4 weeks when there is clinical indication for intervention^3.

Management

The management of pancreatic fluid collection requires a multidisciplinary team approach and is tailored to each individual patient. Conservative management is usually preferred as majority of pancreatic fluid collection are asymptomatic and resolve or reduce in size often with time.

Nutrition started within 48 hours from admission, can reduce risk of infection and in some studies reduces mortality⁶. Enteral nutrition is preferred over parenteral nutrition, preferably oral intake if tolerated and with no difference between nasogastric and nasojejunal feeding only in tolerability⁷. A standard polymeric diet is recommended with enteral feeding and pancreatic enzymes should only be supplemented to those with obvious pancreatic exocrine insufficiency^7. 

Antibiotic is only indicated if there is evidence of infection, and the antibiotic choice should cover gut derived bacteria ³. Multi drug resistance infection is an independent predictor for mortality⁸ hence antibiotics should be used carefully, and the use of prophylactic antibiotics is not recommended⁹.

Invasive intervention is considered in³:

1. Clinically suspected or proven infected necrosis and in persistent organ failure or failure to thrive in those with acute necrotizing pancreatitis despite optimal medical therapy but preferably when the necrosis has walled off.
2. In symptomatic patients with persistent abdominal pain, weight loss and anorexia, gastric outlet syndrome and duodenal or biliary obstruction & should be performed 4-8 weeks after the onset of pancreatitis.

In the setting of pseudocysts, endoscopic drainage is the approach of choice over percutaneous and surgical approach if on EUS assessment the cyst could be drained safely into the stomach or the duodenum as studies have shown it improves quality of life, reduces costs and hospital stay compared to others ¹⁰. Despite conflicting evidence currently LAMS are preferred over plastic stents but comes with an increased costs with studies showing 91-100 % success rate, 90- 100 % resolution of the pseudocyst and on average less than 5% - 10 % complication rate ^11,12. If the pseudocyst communicates with the main pancreatic duct, a transmural approach alone might not result in clinical success and when coupled with transpapillary approach might improve outcomes although there is conflicting evidence and the ESGE guidance advice against the coupled approach ³.

For infected pancreatic necrosis & depending on the location of necrosis /collection, endoscopic or percutaneous drainage should be considered first over surgical intervention. Endoscopic drainage has a high technical and clinical success rate ranging from 82% to100% while percutaneous drainage has around 50% clinical success rate ^12,13. When compared to surgical drainage, endoscopic approach reduces the pro-inflammatory response and proves outcome¹⁴. EUS guided drainage is preferred over the old direct puncture approach (95% vs 35%–66%) and has less complications (0%–4% vs 13%–15%) ¹⁵. EUS guided lumen apposing metal stents (LAMS) have simplified endoscopic drainage when compared to plastic or fully covered metal stents ¹⁶. They have a unique single delivery system and were designed to reduce migration, promote lumen apposition and tract maintenanceand are preferred over plastic stents alone especially if necrosis is > 30% ^15,16.

Multiple direct endoscopic necrosectomies through LAMS are usually needed to achieve response often with insertion of plastic stents to ensure stent patency when endoscopic drainage is the only intervention ¹⁷. There is evidence to suggest that dual modality drainage (both endoscopic and percutaneous drainage) can reduce risk of pancreaticocutaneous fistula ¹⁸. The ESGE & NICE also recommends dual modality drainage in walled off collection especially where the collection extends into the pelvic paracolic gutters ^3,19.

Before deciding on removal of the endoscopic or percutaneous drain, a repeat cross section imagine is usually undertaken to assess the response to the intervention. Percutaneous drainage should be removed when the effluent is clear and is less than 5o ml / hour and with no evidence of pancreaticcutaneous fistula ³.

When LAMS have been used to achieve endoscopic drainage, ESGE guidelines recommends removal in 4 weeks to avoid stent related adverse events such as bleeding ³. However recently, Nayar M et al reported bleeding as 1.9 % which was not associated with the interval between LAMS insertion and removal and with buried stent reported as the most common delayed AE occurring in 4.7 % of patients ²⁰.  Generally, the same team inserting the LAMS should be responsible for removing them. Delayed removal is sometimes necessary in patients with significant pancreatic necrosis with minimal clinical success at 4 weeks.

Conclusion

Management of pancreatic fluid collection requires a multi-disciplinary team approach and should be tailored to each individual patient. Asymptomatic patients do not require any intervention regardless of the size of the peripancreatic collection. If intervention in required; it is preferable to delay for 4 weeks from the onset of pancreatitis to ensure the collection is walled off to avoid complications. Endoscopic approach is preferred and currently LAMS seems to be the preferred stent of choice. Nutritional support is the most important step in management and should be initiated as soon as the patient is admitted.

Author Biographies

Dr Bidour Awad Elkarim, MBBS, MRCP UK, MRCP Gastroenterology, post CCT fellow in pancreatobiliary medicine and hepatology clinical trials, has a special interest in pancreatobiliary medicine and advanced endoscopy, and is currently undertaking advanced training in ERCP and EUS at the Freeman hospital in Newcastle upon Tyne NHS Foundation Trust.

Dr Manu Nayar is a full time Consultant Gastroenterologist at the Freeman Hospital, Newcastle upon Tyne NHS Foundation Trust since 2008. His subspecialty area of interest is hepatobiliary and pancreatic medicine including endoscopic management i.e., therapeutic ERCP and endoscopic ultrasound. He is well published in peer reviewed journals in this area. He has been the director of endoscopy and the training centre lead for Newcastle upon Tyne Hospital. He is the President of the UK & Ireland EUS society. He is a committee member of the pancreatic society of GB & Ireland, the BSG endoscopy section & the European EUS organisation. He organises an annual national course for ERCP and EUS for nurses and is the course director the JAG EUS basic skills course. He is a course director of various national meetings & is a faculty on multiple national live endoscopy events.  He is involved in national trials relating to pancreaticobiliary (PB) endoscopy and is well published in this field.